Cs20

National Conference on Recent Research in Engineering and Technology (NCRRET -2015) International Journal of Advance Engineer ing and Research Development (IJAERD) e-ISSN: 2348 - 4470 , print-ISSN:2348-6406

Review of Various Web Page Ranking Algorithms in Web Structure Mining Asst. prof. Dhwani Dave Co mputer Science and Engineering DJMIT ,Mogar

Abstract: The World Wide Web contains large amount of data. These data is stored in the form of web pages .All these pages can be accessed using search engines. These search engines need to be very efficient as there are large number of Web pages as well as queries are submitted to the search engines. Page ranking algorithms are used by the search engines to present the search results by considering the relevance, importance and content score. Several web mining techniques are used to order them according to the user interest. In this paper such page ranking techniques are discussed.

Keywords: Web Content M ining, Web Usage M ining, Web Structure M ining, PageRank, HITS, Weighted PageRank

I. INT RODUCTION The web is a rich source of information and it continues to increase in size and difficulty.

Efficient and effective retrieval of the necessary web page on the web is becoming a challenge aspect now days [1]. The Web is unstructured data warehouse, which delivers the mass amount of info rmation and also enlarges the complexity of dealing informat ion fro m different perspective of knowledge searchers, business analysts and web service providers [2].

Beside, the Google report on in 2008 that there are 1 trillion unique URLs on the web [3]. Web has grown enormously and the usage of web is unbelievable so it is essential to understand the data structure of web.

Because of the massive amount of information it becomes very hard for the users to find, ext ract, filter or evaluate the relevant information. This issue lifts up the attention to the obligation of some technique that can solve these challenges.

The paper is organized as follows - The categories of Web Mining are discussed in Section 2.

Section 3 exp lains the important of Web Page Ranking and two important algorith ms such as Hypertext Induced Topic Selection (HITS) algorith m and PageRank algorithm. In section 4, we explore the comparison between Web Page Ranking algorith ms used. The Conclusion remarks are g iven in Sect ion 5.

II. WEB MINING CATEGORIES Web Mining consists of three main categories based on the web data used as input in Web Data Mining. (1) Web Content Mining; (2) Web Usage and (3); Web Structure Min ing.

A. Web Content Mining Web content mining is the procedure of retrieving the information fro m the web into more structured forms and indexing the information to retrieve it quickly. It focuses mainly on the structure within a web documents as an inner document level [9].

B. Web Usage Mining Web usage min ing can be defined as one of the application of data min ing techniques to discover interesting usage patterns from web usage data, in order to understand and better serve the needs of web-based applications [2].Web-usage mining mines the secondary data derived from the behavior of users while interacting with the web. This includes data fro m Web server-access logs, pro xy-server logs, browser logs, user profiles, registration data, user sessions or transactions, cookies, bookmark data etc [9].

C. Web Structure Mining Web structure min ing is defined as the process by which we discover the model of link structure of the web pages. We classify the lin ks; generate the ease of use information such as the similarity and relations among them by taking the advantage of hyperlink topology [4]. PageRank and hyperlink analysis fall in this class.

Overview of these three web mining categories is explained and compared in the fo llowing Table 1:

Criteria

Web

Web Mining Content Web Usage

Web

Mining

Mining

View of Data

-Unstructured -Structured

Main Data

- Text documents -Hypertext documents

User Interactivity -Server Logs (log-files) -Browser Logs

Representation

-Ba g of words , n-gra m Terms , -Phrases, Concepts or Ontology -Rela tional Lea rning

Method

-Ma chine Lea rning -Sta tisti cal (i ncluding: NLP)

Application Categories

-Ca tegoriza tion -Clus tering -Finding extra ct rules pa tterns in text

-Relational Table - Graph -User Behavior -Machine Learning -Statistical Association rules -Site Construction -Adaptation and management -Marketing, -User Modeling

Structure Mining -Link Structure -Link Structure

-Graph -Web pages Hits Proprietar y algorithm -Web PageRank

Categoriza tion Clustering

Table 1: Co mparison of different Web Mining categories

III. WEB PA GE RANKING A LGORITHM Page ranking algorith ms are used by the search engines to present the search results by considering the relevance, importance and content score The web min ing techniques are used to order them according to the user interest. Some ranking algorith ms depend only on the analysis of the link structure of the documents i.e. their popularity scores (web structure mining), whereas others look for the actual content in the documents (web content mining), while some use a combination of both i.e. they use content of the document as well as the link structure to assign a rank value for a given document.

There are number of algorith ms proposed based on lin k analysis. Three important algorith ms, such as PageRank, Weighted PageRank and HITS (Hyperlin k Induced Topic Search) are discussed below.

A. PageRank Algorithm (Google) The “PageRank” algorith m, proposed by founders of Google Sergey Brin and Lawrance Page, is one of the most common page ranking algorith ms . This algorith m is also currently used by the leading search engine Google. The algorith m uses the linking (citation) info occurring among the pages as the core metric in ranking procedure. Existence of a link fro m page p1 to p2 may indicate that the author is interested in page. The PageRank metric PR(p), defines the importance of page p to be the sum of the importance of the pages that point to p. More formally, consider pages p1,…,pn, wh ich link to a page pi and let cj be the total number of links going out of page pj. Then, PageRank of page pi is given by: PR(pi)=d+(1-d)[PR(p 1)/c1+…+PR(pn)/cn] where d is the damping factor.

This damping factor d shows that users will only continue clicking on links for a fin ite amount of time before they get distracted and start exploring something completely unrelated. With the remain ing probability (1-d), the user will click on one of the cj lin ks on page pj at rando m. Da mp ing factor is usually set to 0.85. So it is easy to infer that every page distributes 85% of its original PageRan k evenly among all pages to which it points.

Problems of PageRank Algorith m are: It is a static algorith m that, because of its cumulat ive scheme, popular pages tend to stay popular generally. Popularity of a site does not guarantee the desired information to the searcher so relevance factor also needs to be included. It should support personalized search that personal specifications should be met by the search result.

B. HITS (Hyper-link Induced Topic Search) Algorithm (IBM) It is executed at query time, not at indexing time, with the associated hit on performance that accompanies querytime processing. Thus, the hub(going) and authority(coming) scores assigned to a page are queryspecific. It is not common ly used by search engines. It co mputes two scores per document, hub and authority, as opposed to a single score of PageRank. It is processed on a small subset of „relevant‟ documents, not all documents as was the case with PageRan k. Th is algorith m was given by

Kleinberg in 1997. According to this algorithm first step is to collect the root set. That root set hits from the search engine. Then the next step is to construct the base set that includes the entire page that points to that root set. The size should be in between 10005000. Third step is to construct the focused graph that includes graph structure of the base set. It deletes the intrinsic link, (the link between the same do ma ins).

Then it iterat ively co mputes the hub and authority scores.

Mining technique used

PageRank

HITS

Weighted Page Rank

Web Strucutre

Web Structure

Web Structure

Mining

Mining,

Mining

Web Content Mining

Problems of HITS Algorith m are as follow: [5][6] 1 Mutually reinforced relationships between hosts. Sometimes a set of documents on one host point to a single document on a second host, or sometimes a single document on one host points to a set of document on a second host.

2.

Automatically generated links. Web document generated by tools often have links that were inserted by the tool.

3.

Non-relevant nodes. Sometimes pages point to other pages with no relevance to the query topic.

Methodology

pages.

Weight of web page is calculated on the basis of input and outgoing links and on the basis of weight the importanc e of page is decided.

Computes scores at index time.

Functionality

Compute s scores of n highly relevant pages on

Computes weight of the web page at the time of

the fly.

indexing.

High

Middle

High

Black links

Content, Back, forward

Back links and Forward

links

links

- Results are sorted on the importanc e of pages.

IV COMPA RISION Based on the literature analysis, a co mparison of some of various web page ranking algorithms is shown in table 2. Co mparison is done on the basis of some parameters such as main technique use, methodology, input parameter, relevancy, quality of results, importance and limitations.

Criteria Algorithm

It compute s hubs and authority of the relevant

pages.

C. Weighted PageRank Algorithm Wenpu Xing and Ali Ghorbani proposed a Weighted PageRank algorith m which is an extension of the PageRank algorith m. This algorith m assigns a larger rank values to the more important pages rather than dividing the rank value of page evenly among its outgoing linked pages, each outgoing link gets a value proportional to its importance. In this algorithm weight is assigned to both backlin k and forward lin k.

Incoming link is defined as number of link points to that particular page and outgoing link is defined as number of links goes out. This algorithm is more efficient than PageRank algorithm because it uses two parameters i.e. backlink and forward link. The popularity from the number of in links and outlinks is recorded as W in and W out respectively. W in (v, u) is the weight of link (v, u) calculated based on the number of in links of page u and the number of in lin ks of all reference pages of page v. [7][8]

This algorithm computes the score for pages at the time of indexing of the

Accur acy

Input Parameter

High. Back links are considere d.

Importance

Moderat e. Hubs and authoriti es scores are utilized.

O(log N)

Complexity

<O(log

High. The pages are sorted according to the importanc e.

O(log N)

N) Results come at the time of indexing and not at the query time.

Limitation

V.

Topic drift and efficienc y problem.

Relevancy is ignored.

CONCLUSION

In this paper it has been mentioned the introduction of web min ing and its related categories such as web content mining, web structure min ing and web usage min ing and are also tabulated to provide comprision. The main goal of search engines is to provide relevant informat ion to the users to cater to their needs based on the query they provide.

Therefore, finding the relevant content of the Web and retrieving information according the user’s interests and needs have become increasingly important. The different algorith ms used for link analysis like PageRan k (PR), Weighted PageRank (WPR), Hyperlink-Induced Topic Search (HITS) algorith ms are discussed and compared depending on which the aim is to discover an efficient and better system for mining the web topology to identify relevant web pages.

REFERENCES [1] Brin, and L. Page, The Anatomy of a Large Scale Hypertextual Web Search Engine,, Co mputer Network and ISDN Systems, Vol. 30, Issue 1-7, pp.

107-117, 1998.

[2] C. Ding, X. He, P. Husbands, H. Zha, and H Simon, Lin k analysis: Hubs and authorities on the world. Technical report: 47847, 2001.

[3] J. Hou and Y. Zhang, Effect ively Finding Relevant Web Pages from Linkage Information, IEEE Transactions on Knowledge and Data Engineering, Vo l. 15, No. 4, 2003.

[4] Zakaria Su liman Zubi, Marim Aboajela Emsaed.

2010. Sequence min ing in DNA chips data for diagnosing cancer patients.In Proceedings of the 10th WSEAS international conference on Applied computer science (A CS'10), Hamido Fujita and Jun Sasaki (Eds.). World Scientific and Engineering Academy and Society (WSEAS), Stevens Point, Wisconsin, USA, 139-151.

[5] Ashish Jain, Rajeev Sharma, Gireesh Dixit, Varsha Tomar ,”Page Ran king Algorith ms in Web Mining, Limitations of Existing methods and a New Method for Indexing Web Pages”, 2013 IEEE International Conference on Communicat ion Systems and Network Technologies.

[6] Miguel Go mes da Costa, Júnior Zhiguo Gong,” Web Structure Mining: An Introduction”, proceedings of the 2005 IEEE International Conference on Information Acquisition June 27 July 3,2005, Hong Kong and Macau, China [7] Seifedine Kadry , Ali Kalakech ,” On the Improvement of Weighted Page Content Rank”, Journal of Advances in Computer Networks, Vol. 1, No. 2, June 2013.

[8] Rashmi Rani, Vinod Jain ,” Weighted PageRank using the Rank Improvement” International Journal of Scientific and Research Publications, Volu me 3, Issue 7, Ju ly 2013.

[9] Zakaria Su liman Zubi, “Ran king WebPages Using Web Structure Mining Concepts ” Proceeding of the 12th WSEAS international conference of Recent Advances in Teleco mmunicat ions, Signals and Systems, March 2013.